►
From YouTube: Community Meeting 10.26.17
Description
No description was provided for this meeting.
If this is YOUR meeting, an easy way to fix this is to add a description to your video, wherever mtngs.io found it (probably YouTube).
B
B
Okay,
alright,
so
we
didn't
have
a
ton
on
the
agenda
for
today,
basically
just
some
basic
housekeeping
and
announcement,
and
we
also
have
some
questions
coming
in
I
stock.
It
again
I
did
put
out
the
challenge
like
any
any
questions.
You
have.
You've
got
a
full
room
here,
stump
them
and
maybe
you'll
get
something
cool.
So
you
know
there's
always
that
plus
it's
fun
to
stump
the
team.
B
B
If
you
are
headed
to
coupon
or
cloud
native,
comps,
they're,
essentially
the
same
thing
ones
right
next
to
each
other
in
audits,
in
Austin
the
first
week
of
December
we've
got
you
probably
seen
on
the
agenda.
There's
a
couple
of
sessions
that
are
listed
as
GRP
sisig,
which
fun
story
made.
The
kubernetes
community
asked
if
they
had
incorporated
in
g
RPC,
because
the
term
is
so
associated
with
kubernetes.
B
You
know
they
thought.
Perhaps
we
were
now
merging
and
I
had
to
say
like
no,
it's
not
all
about
you
so,
but
so
that
is
a
term
that
we
don't
use,
but
they
do
and
CCF
uses
it
a
lot
too,
but
it
is
for
working
groups,
and
you
know
we
don't
have
the
formal
process
yet
that
will
be
coming,
but
I
know
a
couple
of
people
at
River
East
out.
B
There
was
some
discussion
about
you
know
doing
like
kind
of
a
little
hack
day
or
something
like
that
with
folks
together
and
I
mentioned,
we've
got
the
state.
So
if
you
are
interested
in
doing
something
with
this
space,
you
know
if
you've
got
something
you
want
to
just
get
a
bunch
of
people
together
and
let's
talk
about
something
with
the
project.
Please
reach
out
to
me
and
we'll
work
it
out,
we'll
make
it
happen.
B
D
I
think
I
think
just
a
shot
a
bit,
so
we
are
approaching
the
end
of
the
year,
but
we
will
still
try
to
keep
up
our
six
degrees.
It
is
one
of
the
things
that
you
might
notice
is
that
our
arty
releases
may
stay
our
seat
for
longer
periods,
then
than
usual,
because
it
will
just
take
a
longer
time
for,
like
other
teams
and
others
and
people
to
verify
or
RC
packages.
D
So
so,
as
we
do
our
next
maybe
couple
of
releases
and
the
holidays
when
we
probably
are
see
releases,
if
somebody
in
the
community
can
actually
test
validate
those
RC
packages,
provide
us
feedback
that
will
actually
help
us
improve
the
quality
of
these
releases
and
get
us
out
quicker
as
well.
So
just
wanted
to
let
everyone
know
what's
happening
for
the
next
couple
of
weeks.
B
E
B
D
Yes,
so
so
basically,
I
think
because
I
don't
think
we
package
the
FEC
core
separately.
So
there's
no
specific
RC
for
that,
but
there
is
branch
is
available
and
the
packages
have
started
rolling
out,
like
I.
Think
many
of
the
rap
languages
like
Python
and
so
on
are
ready
to
cut.
So
you
should
start
seeing
seeing
the
RC
packages
go
into
1.7
a
sooner
they
go
like,
but
in
terms
of
tagging
it
your
time
wants.
One
we
are
most
of
the
exact
agents
are
verified,
which
would
be
any
day
now.
Action,
juice.
D
B
G
A
G
G
G
G
Proxies
if
ever
I
thought
the
traffic
was
within
the
cluster,
but
our
our
kind
of
problem
is
like
right.
Now:
we've
got
a
couple
and
a
rumble
set
and
a
couple
of
leg
load
balance
services.
So
we've
got
a
few
public
ip's
that
we're
trying
to
kind
of
balance
across
and
it
doesn't
seem
like
it's
a
super
performance
solution.
I'm
doing
this,
like
a
healthcare
field
for
load
balance
thing
below
necessarily
evenly
spread.
We
don't
have
a
great
visibility
into
you
know
where
traffic
is
going.
G
Latency
is
coming
from
because
there's
so
many
hops,
but
kind
of
she's
you'll
be
too
cute
proxy
to
the
pods.
So
I'm
wondering
if
you
guys
have
advice
about
I,
guess
configuring
that
you
know
a
deserve
ability
for
that,
come
in
the
wild
and
if
something
like
a
Honda
boy
is
in
the
l7,
we
don't
we
don't
have,
but
we've
only
got
engine
access
and
egress
I.
Don't
think
it
supports
Jer
BC
right
now,
but
if
adding
like
an
envoy
ingress,
it
would
be
the
best
thing
to
do
forward.
A
A
F
F
F
F
F
A
F
G
Yeah
I
guess
corollary
thing
here
is
I:
haven't,
had
a
ton
of
success
with
kind
of
understanding,
our
clients
are
Secord,
client
and
the
server
is
generous
ego,
but
in
both
cases,
I've
kind
of
had
trouble
getting
like
her
request.
Tracing
of
the
latency
one
foot
in
the
GTR,
PC
stack,
and
so
we've
tried
kind
of
as
best
as
we
can
to
eliminate
those
as
like
a
potential
source
of
latency.
G
Something,
like
you
know,
the
cube
on
the
client
side
of
something
but
I,
don't
know
if
there's,
if
there's
a
document
or
something
you
guys
could
refer
for
me
to
to
get
kind
of
better.
You
know:
we've
gone
like
stock
driver
tracing
through
the
whole
thing,
but
the
GRDC
chunk
is
kind
of
a
black
box
right
now,.
B
F
H
G
A
G
It
doesn't
client-side
or
server-side,
we
don't
see
any
of
the
jaribg
internals,
so
we
don't
know
or
like
client-side
queuing
or
like
we
can't.
We
can't
account
for
that
right.
We
just
see.
Sometimes
the
trpc2
all
goes
into
the
juror
busy
client
and
it
comes
out
150
milliseconds
later,
sometimes
it
comes
out
50
milliseconds
later
sometimes
it
comes
out.
You
know
like
it
times
out
and
adjust
the
color,
so
any
anything
to
get
more
visibility
into.
That
would
be
amazing,
so.
G
H
So
I
think
that
might
give
you
the
information
that
you're
looking
for
that
has
essentially
callbacks
that
it
will
invoke
through,
like
the
lifecycle
of
the
connection
and
then,
and
so
you
can
collect
timing
information
from
that
I,
don't
think,
there's
any
sort
of
out
of
the
box.
You
know
thing
that
we
have
that
will
take
all
the
information
and
presented
nicely
to
you.
You
have
to
sort
of
fill
that
your
cellphone's
off
of
the
call
to
X
that
we
provide
I.
G
Mean
I,
suppose
I
guess
the
ask
like
the
higher
level
ask
would
be
you
know:
it'd
be
cool
to
see
some
kind
of
like
reference.
I
know,
there's
the
load-balancing
dog,
but
you
know
understanding
a
little
accessing
to
like
a
GPC
service.
I,
don't
know
if
I'm
like
I
said
I,
don't
know
folks
again
of
examples
of
that
we
were
a.
D
H
F
D
F
F
G
And
just
to
be
clear,
the
thing
that
I
raised
here
I
think
is
subtly
different
from
the
max
connections,
because
you
know
we
do
have
that
and
go,
but
we
could
potentially
like
what
would
be.
Ideally
thing
is
if
we
could
have
potentially
many
idle
connections
but
constrain
the
the
number
of
go
routines
that
I
believe
you
are
working
on
extremes
at
any
point.
H
G
G
G
H
A
H
I
believed-
and
so
you
could
at
that
point
essentially
just
increment
a
counter
or
or
your
ticket
counter
and
and
then
you
can
actually
reject.
You
can
return
an
error
to
reject
the
stream
I,
don't
know
how
did
we
intend
to
support
that
forever,
like
it's
like
a
good
long-term
solution,
I
think
the
other
languages
don't
have
an
analog
and
I.
Don't
I,
don't
even
know
that
facility
is
a
good
one
to
be
using,
but
it
definitely
works
for
now.
I
think
it
says
it's
experimental
in
the
comments
yeah.
F
I
think
people
sauce
is
something
where
you
can
delay
accepting
or
you
can
be
told
you
can
sort
of
how
things
sit
for
a
little
bit
before
you
accept
the
stream
I
think
Java,
which
I
mean
that
is
a
specific
API
I,
think
that
you
have
to
use
in
Java.
You
have
to
use
some
like
interceptor,
it's
something
that
way,
maybe
possible
across
languages,
but
it's
not
you.
H
F
H
You
could
use
a
server-side
interceptor
also
to
accomplish
and
that
we've
definitely
been
more
stable,
essentially
to
return
an
error
back
from
the
from
the
intercepted
app
and
prevented
from
reaching
the
application
code.
So
he'd
still
incur
the
cost
of
you
know,
processing
the
headers
and
and
all
of
that,
but
at
least
seen
any
chilled
your
application
code
from
being
executed.
G
G
C
G
H
That
sounds
like
a
good
idea:
I
I
wonder
about
something
something
like
that
into
x-men
http/2,
because
either
it
would
have
to
be
global
to
your
binary
and
you
may
land
multiple
servers
and
you
want
to
apply
different
limits
across
them,
or
we
essentially
have
to
be
for
connection,
which
is
you
know
what
we
already
have.
The
controls
for,
and
so
I
don't
know
if
they
would
be
accepting
of
something
hopeful
and
further
connection,
wouldn't
wouldn't
help
us
on
this.
So
yeah,
where
we
can
talk
about
this
would
probably
be
a
good
idea.
Yeah.
H
F
What's
for
configuration
might
need
to
tweak
I
do
think
that
we've
got
less
practice
doing
this,
but
you
would
probably
want
at
least
a
max
age
and
keep
alive,
although
keep
alive
has
point
two
points.
So
if
you
end
up
flipping
into
l7,
foxy
is
in
the
middle
that
proxy
we
need
to
participate.
We
need
to
do
keep
alive
to
defy
it
and
then
IMAX
streams
per
client
that
you
already
have
BEP
is
good
for
any
environment.
That's
that's!
Not
a
concern,
so
I
think
that
that's
that's
much
simpler
sort
of
answer.
G
C
C
Install
every
dependency
on
my
machine
and
it
will
have
to
worry
about
every
version,
so
I
started,
throwing
a
bunch
of
stuff
in
docker
containers
and
I
wasn't
curious.
If
there's
for
this,
and
if
there
was
I
can
work
on
porting
that
into
the
open
source.
Otherwise
I
was
just
thinking
with
those
source.
For
now.
F
F
These
containers
would
also
probably
be
set
up
so
that
whenever
you
just
run
it,
it
probably
just
runs
the
proto
C
plug-in
just
unconditionally
as
the
entry
point
into
the
doctor
yeah.
But
you
can
have
a
pretty
slick
experience
if
you're
in
the
doctor
world
that
make
sense,
this
should
be
really
trivial
to
to
maintain
as
well
yeah
they're.
D
I
Can
you
hear
me
yep
yeah,
just
had
a
good
question
about
that.
Cocoapods
update
I
to
fix
was
to
run
a
previous
version
of
cocoa.
Pods
is
just
forgiving
it
that
was
so
the
case
or
if
someone
had
looked
at
it
over
there
I've
not
had
a
chance
to
test
it,
because
we
didn't
want
to
push
all
the
stuff
to
the
newer
version
and
have
error
again,
but
are
we
just
sticking
to
the
1.3.1
on
that
or
just
what?
What's,
the
update
Thanks.
I
A
A
B
B
B
E
So
I
can
take
a
crack,
so
census,
veena
library
revealed
internally
at
Google
when
we
bake
GRCC,
we
wanted
to
use
package,
and
so
there
is
another
team
at
Google
census
team.
They
are
happening
to
build.
Library
are
working
on
open
substances,
libraries
and
from
what
I
know
their
plan
is
to
make
it
fully
compatible
in
formats
with.
E
B
Have
some
good
news
that
I
can
share,
which
is
that
someone
that
I
was
working
with
on
the
developer
relations
team
on
the
both
side
has
now
moved
over
a
fence
aside
and
I
made
it
very
clear
to
her
as
soon
as
she
was
full
up
to
speed,
I
was
bringing
her
in
here.
She
was
going
to
get
it
and
take
all
census
questions
because.